Google launches specialised eighth-generation TPU to welcome the Agentic AI era

Las Vegas (ANTARA) - Google Cloud has officially introduced the eighth generation of Tensor Processing Units (TPU), namely TPU 8t and TPU 8i, designed to accelerate artificial intelligence (AI) workloads in the Agentic era.

SVP and Chief Technologist for AI and Infrastructure at Google Cloud, Amin Vahdat, stated at the event The Future of AI Infrastructure, Google Cloud Next ’26, in Las Vegas, United States, on Tuesday, that this year, two specialised chips are being launched simultaneously to welcome the Agentic AI era.

“For the first time, two custom-made TPUs, 8t and 8i, have been entirely designed from scratch to meet current needs. TPU 8t is for training. This is the main engine. This is the largest scale. TPU 8i is truly built for inference,” said Vahdat.

Vahdat explained that for years, TPUs have supported leading foundational models, including Gemini. The combination of these eighth-generation TPUs will deliver scale, efficiency, and capabilities encompassing training workloads, serving, and agentic needs.

Therefore, 2026 marks a pivotal moment for the technology giant to introduce two highly powerful specialised chips simultaneously, the result of the past two years of development.

Vahdat elaborated that TPU 8t is optimised for training and capable of running even the most complex models within a single massive integrated memory pool.

Meanwhile, TPU 8i is specifically designed to enable AI agents to complete processes extremely quickly, providing a satisfying user experience.

Vahdat emphasised that both chips are purpose-built from scratch for training and inference efforts, not mere derivatives of one another. Specifications, capabilities, connectivity—all are designed differently due to their specialisation for individual needs.

“If you need main power for training, you have 8t. If you need lightning-fast inference, you have 8i,” Vahdat clarified.

Through the commercial availability of these latest chips by the end of the year, Google Cloud aims to redefine the boundaries of AI capabilities for developers and global companies.

The utilisation of this technology, including in Gemini Enterprise services, is expected to accelerate scientific discoveries to transformations in the healthcare sector on an international scale.